Introducing Open WebUI for Local LLMs
Many of us don't really know how to install large language models or AI models locally. You've seen LM Studio's text generation web UI, but today I'd like to showcase a newly updated UI called Open WebUI, or previously known as Web UI Ollama. Open WebUI is an extendable, user-friendly, self-hosted web UI, and it's designed to operate entirely offline and 100% secure. It supports various large language model runners, including Ollama and OpenAI compatible APIs.
Now, throughout today's video, we're going to be showcasing how you can install it locally and showcase what you can actually do with it. So with that thought guys, stay tuned and let's get straight into the video.
Installing Pinocchio for Simplified AI Setup
So let's just get started. So you can actually install this using Docker, which is an easy way to install it, but there's also another easy way, which is utilizing Pinocchio. This is a tool that lets you install almost any AI model easily and locally on your computer. You just need to install their Pinocchio application, which lets you install all these other apps. Throughout its app, you can install AI Town, L-Chat, and so many other models. You can even install web UI, stable diffusion models, and such forth.
So with that thought guys, just go over to the Pinocchio website, which I'll leave a link to in the description below. Click on this download button, it'll take you to the docs. You can choose whatever operating system you want to utilize. So if you have a Windows, make sure you install Pinocchio for Windows. If you have a Mac, you make sure you install from Mac and just follow through with the installation.
Now, just to showcase how you install from Windows, just click on the download button over here. It will then download a zip file. You just need to unzip it, then you need to just click on the install file and it will then prompt up a Windows protected PC. But they state that this message shows up because the app was downloaded from the web, and this shows up for Windows apps, or Windows does this for apps downloaded from the web. So you just need to click on more information, run anyway. And once that is done, you can then open up Pinocchio. It is easy as that, guys. and once that is done, you're basically halfway there.
Downloading and Installing Open WebUI with Pinocchio
You will then be prompted up to this screen over here. You just need to simply click on visit Discover page. Once that is done, we need to head over to Open WebUI. You can just simply search it up, but you can even paste the Git URL to easily install it. In this case, we're just going to open up Open WebUI. We're going to click on download. You can name it as whatever you want or save it as whatever you want, but in this case, we're just going to save it as the base default name. And it's going to start installing it all by itself, automatically and autonomously. We don't need to intervene, we just need to let the application install it by itself. So once it finishes installing, I'll be right back.
And there we go. We have it over here on our dashboard where we can see Open WebUI fully installed. It's currently showing not running because we don't have it opened up, but if we want to, we can just simply click on it and we can then install it. And once this is done, we can then have it so that it can install these various components such as Git, zip, you have Conda, registry as well as CUDA. So you can just simply click on install. It's going to take a bit longer to do this, so once it's finished, I'll be right back.
Handling Dependencies: Installing Ollama
And there we go, we have it installed. It'll pop up this window which says install complete. you can then click okay and then proceed to start it up. So I actually had to install a couple other packages afterwards, so once that was actually completed, I had to just go back to the homepage, go click on the Open WebUI app that we just installed, and then you can see that you can start it up by clicking this button over here.
Now, once this is done, it will start on focusing on starting it up, but it looks like we're going to need to launch Ollama before proceeding. So we need to install Ollama. Once that is done, we can then have it running. So I'm going to install it for my Windows OS, so I'm going to quickly install this and get right back to you.
Sorry for being repetitive, but this month we had insane partnerships with big companies giving out subscriptions to AI tools completely for free. These are tools that will streamline your business's growth and improve your efficiency. Just being a Patreon this past month, you were given access to six paid subscriptions completely for free.
Not only do you access these subscriptions, but you gain the ability for consulting, networking, collaborating with the community as well as with myself. You get access to daily AI news, resources, giveaways, and so much more. If you're interested, check out the Patreon link in the description below to gain access to these benefits.
Final Setup and Launching Open WebUI
So once you have installed the install file for Ollama, you just need to open it up. There will be an install button, you just simply click on it and we can see it's going to start extracting the files from it and install the Ollama model onto our computer. So once this is done, we can then head over back to Pinocchio and this is where we can just simply click on next and start this up quite easily.
So now that we have Ollama installed, you can just simply click on next and it's going to start up Open WebUI. It should take a couple of seconds since this is our first time starting it up, but then we're going to be able to have it fully hosted on our Local Host. So it's going to prompt when it's ready to be opened up on your Local Host, and there we go. It gave us the URL for our Open WebUI, which is our local host over here.
First Look: Navigating the Open WebUI Interface
And we can then start chatting with it. So, you're going to need to actually sign in with Open WebUI, but if you do not have an account, you can just simply click on sign up. And there we go, we can start playing around with the web UI.
And you can head over to the Modelfiles where you can upload your own model files. You can add different prompt templates or even documents so that you can even chat with it. Now if you click on this bottom left profile tab, you can see that there's a settings tab. You can have your chats archived, you can play around with different chatbots, and you can head over to the admin panel where you can even have it connected to a Cloud Server, so other people can integrate or interact with your server or basically your server files that you have uploaded. So this way, whatever chats or whatever documents you have locally installed are uploaded to a private server so other people can actually access it and chat privately. This could be used for companies or people who want to actually interact with whatever files you're working with.
Managing and Adding New LLMs
But let's head back over to the new chat, and we're going to be playing around by adding another model. So you can see on the right-hand side, there's a settings tab, and you can head over to the models tab. It's going to take a couple of seconds to load this up, but we're going to be able to see that you can import different models from Ollama. And to do this, you just need to head over to the Ollama website, which I'll leave a link to in the description below. Head over to the library or you can click on Models Tab, and you can see that there's various models that you can integrate into this.
So you can input Code Gemma, you can have Llama models, LLaVA models, Qwen models, Dolphin models, as well as many other models that are uploaded and compatible with Ollama. So in this case, if you would want to install something like Phi, which is the latest Microsoft model, but under the Phi family—I think this is Phi 2 at the moment—but you can actually copy the link from this. You can head over back to Web UI and you can import that model. And you can see that you can even delete a model that you don't want to work with. You can manage LM, light LM models, and you can even experiment with other models such as uploading your own GGUF model.
So this gives you a lot of flexibility and it allows you to do various different things. Now, the amazing thing with Open WebUI is that they have associated their application with so many amazing features. You have access to GGUF models, you have local RAG integration where you're going to be able to utilize the RAG framework or, sorry, algorithm, which is going to support within the chat experience that you have. And you're going to be able to integrate various other files that you can use RAG to interact with documents.
You have code syntax highlights where it's going to be able to focus on enhanced code readability. You have full Markdown support, you also have voice input support as well as fine-tuned control and advanced parameters. There's so many different things to this and there's various things that I truly recommend that you read through within their GitHub repo because it showcases different sorts of avenues as to what you can do with it.
Conclusion and Final Thoughts
And that basically concludes today's video on Open WebUI. This is a great way for you to host various large language models locally on your computer, 100% secure and private. So if you enjoyed this video, make sure you like it, make sure you turn on the notification bell. And if you haven't already, make sure you check out the Patreon page because this is a great way for you to access different subscriptions completely for free.
Make sure you take a look at the Twitter page. This is a great way for you to stay up-to-date with the latest AI news. And lastly, make sure you guys subscribe, turn on notification bell, like this video, and check out our previous videos so you can stay up to date with the latest AI news. But with that thought guys, thank you guys so much for watching, have an amazing day, spread positivity, and I'll see you guys fairly shortly. Peace out fellas.
